2018 IEEE/ACM International Conference on Utility and Cloud Computing Companion (UCC Companion)

Building Trustable Remote Monitoring and

Management Systems
Maghsoud Morshedi Josef Noll Raheleh Kari
University of Oslo and EyeNetworks University of Oslo Norwegian University of Science and Technology
Oslo, Norway Oslo, Norway Ålesund, Norway
Email: [email protected] Email: [email protected] Email: [email protected]

Abstract—Internet of Things (IoT) is an emerging technology consumption, health checks, measured data of sensors by
that expands wireless and mobile networks into heterogeneous means of self-reporting or installing an agent on remote
network of connected devices. Trustable remote monitoring and devices. In remote management, service providers administer
management systems are required to establish a controlled
environment for new services and devices in order to (i) improve remote devices to perform certain tasks such as software
the quality of existing services and (ii) enable novel services. updates (e.g. patches, firmware updates and configuration
However, monitoring and remote management can cause security changes), disable or enable specific services or functionalities,
and privacy concerns and thus affect the trust formation between reboot or shut down the device, etc.
customer and service provider. This paper introduces a trust The remote monitoring and management have many appli-
model considering institutions as mediators to assess trustability
of remote monitoring and management systems. The proposed cations in medical, industry, home, energy, and transportation.
model considers governance as an approach to audit remote For example, health centers can optimize the use of health care
monitoring and management systems and accordingly provides resources by means of remote management of heart failure
institutional assurance in form of certificate or labels in order using implantable electronic devices, which downloads sensors
to facilitate trust decision making and motivate trustworthy data from patients’ implanted devices [2], [3]. In case of
behaviors. The proposed model utilized the multi-metric method
to measure governance criteria objectively and represent level automotive, the remote monitoring and management system
of trustworthiness with A-F labels. Representing governance can diagnose engine problems or perform health check as well
criteria with labels accompanied by color coding facilitates trust as remote software update of the internal computer.
decision making based on application context or requirements Although there are various technology solutions to im-
for everyone regardless of level of expertise. Meanwhile, issuing prove the security and privacy measures comprising proper
trustworthiness certificate or A-F labels will encourage service
providers to improve trustability of their remote monitoring encryption, authentication, access control, etc [4], [5], remote
and management approaches, which improve acceptability and monitoring and management of device behavior have been
efficiency of managed services. announced as a security strategy. However, remote monitoring
Index Terms—trust, cloud computing, remote monitoring and and management may raise security and privacy concerns
management, IoT, wireless networks, privacy, security. in respect with unregulated monitoring and management of
services devices and accordingly users become reluctant to
I. I NTRODUCTION use managed devices or enable remote monitoring and man-
The Internet of Things (IoT) is emerging and extends the agement on their services.
IT networks with many new connected devices. The number However, the key challenge is to build a trustable remote
of connected IoT devices will grow from 27 billion devices monitoring and management system in order to assure users
in 2017 to nearly 127 billion connected devices in 2030 [1] regarding their security and privacy. In the absence of trust,
in various domains such as ehealth, industry applications, en- users will be reluctant to use or enable remote monitoring
tertainment and transportation. However, the deployed devices and management services due to growing security and privacy
can malfunction and cannot do what they are intended due to breaches. The present paper presents existing trust issues in
technical problems or malicious activities. In the traditional monitoring and management systems and defines objectives in
approach, technicians deployed to customer premises to fix order to build trustable remote monitoring and management
issues, which is a time-consuming and costly approach and it systems by motivating trustworthy behaviors from services
cannot scale to billions of devices. Hence, service providers providers’ side.
have been utilized cloud-based remote monitoring and man- The paper introduces trust issues that may raise security
agement of services and devices. and privacy concerns due to implementation problem or lack
The remote monitoring and management is a process of of information. Thus, developing trust objectives consider
supervising and administration of information systems such expected behaviors to encourage service providers to improve
as network devices, servers, mobile devices, and sensors. trustability of their remote monitoring and management system
In remote monitoring, service providers enable the endpoint as well as facilitate trust formation for users to accept and
devices to report their operating information such as resource actively use managed services or devices.

978-1-7281-0359-4/18/$31.00 ©2018 IEEE 213

DOI 10.1109/UCC-Companion.2018.00059
Authorized licensed use limited to: ULAKBIM UASL - GAZI UNIV. Downloaded on April 25,2024 at 18:57:55 UTC from IEEE Xplore. Restrictions apply.
 Developing trust objectives leads to a trust model, which
can boost the acceptability of remote monitoring and man-
agement systems. The proposed model considers institutional-
based trustability assessment in form of certificates or labeling
program in order to motivate trustworthy behaviors as well as
facilitates trust formation for end-users and improve accept-
ability of new managed services. Hence, the proposed trust
model utilized the multi-metric method to quantify governance
criteria defined for trustability assessment and accordingly
represent assessment with certificate and labels. The trust
labeling program used in the proposed trust model motivates
service providers to improve trustability of their remote mon-
itoring and management approach, meanwhile expedites the
trust decision making for end-users. In effect, trustable remote
monitoring and management system will improve efficiency
and quality of service in terms of performance, cost and user
satisfaction.
The rest of paper is organized as follows: The definition of
Fig. 1. Remote monitoring and management variables and actors in various
trust, distrust, trustor, and trustee presented in Section II. The use cases.
existing trust issues that may raise user security and privacy
concerns appear in Section III. The trust objectives that aim
to improve trustworthiness, as well as proposed trust model, III. T RUST I SSUES
appear in Section IV, while trust labeling description appears
in Section V. Section VI discusses the objectives and trust In each trust relationship, trust issues can affect the trust
model presented in the previous section. Section VII presents negatively and result in distrust. Identifying trust issues helps
related trust models and trust assessment approaches. Finally, to regulate certain actions such that a trustee demonstrates
Section VIII presents research conclusion and future work. proper behavior in a trust relationship. Trust issues have been
studied in various domains such as wireless sensor networks
II. T RUST D EFINITION [11], vehicular communication [12], cloud computing [13],
and software-defined networking [14]. Although remote mon-
In dictionaries, trust is defined as “solid belief in the itoring and management have common trust issues with other
reliability, truth, or ability of someone or something”. The technologies, its applications in various domains and dis-
definition of trust in computer science can be categorized in tributed cloud-based architecture pose additional trust issues.
term of user trust and system trust [6]. The user trust has its Trust issues can be categorized into monitoring and configu-
root in psychology and sociology, with a common definition ration attributes and features, access control, technology and
as “a subjective expectation as an entity has about another’s law enforcement groups. Figure 1 illustrates various use cases
future behavior” [7]. This implies that trust is perceived of remote monitoring and management while different parties
personally. On the other hand, the system trust is a security act as trustee in different use cases. There are various actors in
perspective and defined as “the expectation that a device or remote monitoring and management system that are not trustee
system will faithfully behave in a particular manner to fulfill its while they can access user data and possibly manage remote
intended purpose” [6], [8]. For example, a remote management devices. The behaviors of these actors can directly influence
system is trustworthy if it is reliable to perform as expected, user trust. The following describes trust issues in various use
such that it can manage remote devices to ensure remote cases in order to underline the severity of remote monitoring
devices are in the desired state exactly the same way as it and management issues in different domains.
expected. Therefore, trust can be defined in technology domain The monitoring and management attributes and features
particularly monitoring and management systems as allowing can become the main cause of trustability concerns such
other parties to access vulnerabilities based on their positive that monitoring attribute containing identifiable information or
intention and behavior. In contrast, distrust can be defined as managing features that are not related to quality improvement
avoiding other parties to access vulnerabilities because of their can result in severe security and privacy breaches and accord-
doubtful intentions and behavior. ingly user distrust. For example, monitoring driving habits or
A trust relationship has two actors including trustee and remote management of car functionalities such as brakes or
trustor. The trustor is an actor who trusts the target entity steering can raise trust issues because if it is manageable then
called trustee [9], [10]. For example, a customer who buys there is a possibility to be hacked and misused. Zhang et al.
a service or device is a trustor and the service provider who demonstrated that electronic controller units (ECUs) of cars
performs remote monitoring and management is a trustee in a can be compromised using remote wireless connection so that
system perspective. they could control in-car meter display, car’s lock and steering

214

Authorized licensed use limited to: ULAKBIM UASL - GAZI UNIV. Downloaded on April 25,2024 at 18:57:55 UTC from IEEE Xplore. Restrictions apply.
system by sending control messages to car’s controller area security of user data and managed device. Indeed, monitoring
network (CAN) [15]. and management system can locate in various region in the
Service providers behavior can cause trust issues such that world, so that if a sensor or a service has provisioned on
service provider (trustee) may exploit the vulnerabilities of monitoring and management system of another region then it
trustor in a relationship. The service provider can exploit can raise distrust even though service provider locates in a
monitored data for marketing purposes, without user permis- country that the sensor is being used.
sion. Indeed, a user engages in a trusting action (e.g. enabling
monitoring of a service), when the user realizes a gain (e.g. IV. T RUST O BJECTIVES
improved quality of service) and the trustee does not exploit The trust issues can impede trustors enabling remote mon-
user’s vulnerabilities (e.g. does not exploit monitored data for itoring and management even though trustors would realize
marketing purposes). There are certain trustee’s actions, which a gain so that trustee can experience business loss. Hence,
result in trust issues and accordingly cause distrust across users trust objectives state desired achievements for both trustor
of a service. For example, a service provider can monitor data and trustee in a relationship in order to maximize gain for
that are not related to the quality of service (e.g. location of both sides of a relationship. Trustee (e.g. service provider)
device or identity of the owner of a sensor or device), thus can carry out certain behaviors or practices during information
monitored data can reveal sensitive information. Likewise, the exchange, which improve quality of service and promote
service provider may share monitored data of a service with trust among trustors (e.g. customers). The trustee should
third parties or partners in order to gain benefits rather than only monitor minimized parameters rather than any data or
improving the quality of service. In this case, users may avoid information relating to identifiable individual or device, whilst
managed services even though they realize a gain by using the data is the raw material for improving quality of service.
managed services (e.g. patients may avoid using managed The same applies to the remote management such that trustee
implantable electronic devices). should only configure minimized functionalities rather than
Another doubtful behavior is that a service provider may functionalities that compromise the device such as disabling
delegate the monitoring and management to third-party com- security mechanism remotely. In addition, security and re-
panies, while they do not define responsibilities and policies liability of remote monitoring and management technology
about data handling so that this may cause distrust. For increase trustability of the system.
example, a customer buys a sensor from vendor X, but The trustor will either trust the trustee or withdraw from a
the sensor communicates with a third-party company Y for relationship. In contrary, the trustee may lack the incentive
remote monitoring and management features such as quality of to fulfill trustworthy action and decide to exploit trustors’
service or software updates. Indeed, delegating monitoring and vulnerabilities. A trust objective is to motivate trustee to
management without determining the responsibilities, access carry out trustworthy actions. For example, motivate service
control and data handling policies can establish distrust among provider (trustee) to implement proper security and privacy
customers. measures for their monitoring and management system by
Security and privacy concerns regarding remote monitor- design. Hence, third-party institutional assurance such as cer-
ing and management technology can also cause distrust. In tificates or labeling programs (e.g. from governmental institu-
monitoring and management system, lack of data transmission tions) can motivate trustee (e.g. service providers) to carry out
confidentiality in addition to low level of security may result trustworthy actions. The certificate program presents a binary
in information leakage or unauthorized access to remote de- outcome of trust assessment, whereas labeling program can be
vices or sensors. For example, implementation of CPE WAN represented by granular discrete labels such as A-F alphabet-
management protocol (CWMP) without security mechanism ical labels, which also helps trustors to make a decision more
simplifies communication eavesdropping or performing unau- easily in various situations. Institutions can provide labeling
thorized remote procedure calls (RPCs) in the remote device system such that service providers receive trustworthiness
(e.g. home Internet gateways). The CWMP is not inherently in- label for their remote monitoring and management approaches
secure, but the improper implementation of TR-069 clients and rather than service-based trust assessment. Assessing service
servers may expose vulnerabilities, which affect many devices. providers trustworthiness minimizes overhead of governance
The improper implementation of CWMP practices comprises in service development.
using HTTP instead of HTTPS, lack of data validation [16] on The trustor can trust the trustee if trustee complies with
the parameters used in a configuration, which results in code certain regulations and policies. Hence, establishing remote
injection or lack of proper authentication. monitoring and management governance as a trust objective
The monitoring and management system can be a geograph- will increase the confidence of trustor to participate in remote
ically distributed system so that law and the levels of law monitoring and management. For example, the remote moni-
enforcement that the system operates can cause trustability toring and management governance can comprise auditability
concerns. For example, remote monitoring and management and accountability of trustee regarding monitored data and
system can be geographically distributed and users monitored managed devices and services. In this regard, (i) the purpose
data can be stored in another country. In this case, which of data collection (monitoring) and data processing should be
law or levels of law enforcement will assure the privacy and declared (ii) data backup and destruction plan for all storage

215

Authorized licensed use limited to: ULAKBIM UASL - GAZI UNIV. Downloaded on April 25,2024 at 18:57:55 UTC from IEEE Xplore. Restrictions apply.
 itoring and management systems. However, result of gov-
ernance can be represented with a certificate or labels in
order to facilitate trust formation. The governance can be
done by observation or self-declaration approach, which this
paper considers self-declaration approach in order to reduce
governance overhead for implementing certificate or labeling
program. On the other hand, the collected information by
governance has to be measurable so that information can be
represented by certificate or labels in an automated process.
This paper introduces generic and high-level governance
criteria for remote monitoring and management systems by
performing a case study in a company, which performs remote
monitoring and management of Wi-Fi access points. How-
ever, the introduced governance criteria can be generalized
to different domains as well. The criteria were selected in a
Fig. 2. Proposed trust model for remote monitoring and management systems. way to identify trust issues in remote monitoring and man-
agement system so that criteria reflect trust issues presented
in preceding section. In addition, the governance criteria can
location is necessary (iii) the physical location of data need to
be extracted from European general data protection regulation
be recorded from law enforcement perspective (iv) data access
(GDPR) [17] as well. Hence, governance criteria provide mea-
policies should be performed in order to identify which data is
sures to identify trust issues in monitoring and measurement,
accessible for partners or third-parties (v) how sensitive data
which in the simple form can consist of yes/no questions.
is safeguarded in storage systems e.g. encryption mechanisms.
However, governance criteria can consider application-specific
The remote monitoring and management are generally im-
details, which may impede generalization to new services and
plemented as a system of systems so that it is part of a
applications and accordingly challenge the whole certificate
bigger system with various types of interconnections. Al-
or labeling program. Table I presents generic and high-level
though actors’ interactions may differ in various use cases,
governance criteria such that service providers can answer
high-level architecture of remote monitoring and management
questions in form of self-declaration in order to allow trust
approach will remain identical. In a remote monitoring and
model to identify existing trust issues in remote monitoring
management system, identifying roles of actors and their
and management approach, regardless of application or service
interactions will help to recognize trust issues and form
type. The criteria categorized to four groups of trust issues
proper trust objective accordingly. This paper introduces a
discussed earlier, which can be extended per domain. Indeed,
trust model, which considers institutional entities as mediators
each category can introduce detailed sub-criteria in order to
in order to facilitate trust formation and motivate trustworthy
provide granular identification of trust issues. Therefore, each
behaviors among trustees. In the proposed model, institutional
category presents sub-criteria one level further in order to
entities assess trustability of trustees’ (e.g. service providers)
demonstrate the process of providing detailed information
remote monitoring and management approach by means of
for certificate or labeling program. Nonetheless, existing sub-
governance criteria. This will also help trustors (e.g. users)
criteria can be detailed into lower level information e.g.
to make a trust decision with confidence and accept trustable
encryption algorithms used, cloud credential policies or en-
remote monitoring and management services rather than reject
cryption key sizes sub-criteria.
services. Figure 2 illustrates the proposed trust model in
order to ensure trustworthiness of service providers remote The governance provides information about trustee’s be-
monitoring and management approaches rather than individual havior and existing trust issues in the system. However,
service assessment. In the model, institutions as a mediators mapping governance to a labeling program, which every-
audit service provider’s remote monitoring and management one can understand easily requires quantifying governance
approach by means of governance criteria so that service criteria. When the governance outcome became measurable,
providers receive trustworthiness certificate or A-F labeling. thus it would be straightforward to issue certificate or label
The certificate or A-F trustworthiness labeling will motivate for service provider’s remote monitoring and management
service providers to improve trustability of remote monitoring approach. Hence, this paper utilizes the multi-metric method
and management approaches, meantime will simplify decision presented by [18] and [19] in order to quantify governance
making for end-user for accepting managed services or de- criteria and accordingly issue certificate or label for trustee’s
vices. remote monitoring and management approach. The multi-
metric method utilizes two parameters: score (Xi ) and weight
V. T RUST L ABELING P ROGRAM (Wi ) or importance of each criterion in order to calculate
The trust model proposed in preceding section introduced overall trustability score of an approach or a system. Therefore,
the essence of governance in building trustable remote mon- overall trustability score can be calculated by a mean square

216

Authorized licensed use limited to: ULAKBIM UASL - GAZI UNIV. Downloaded on April 25,2024 at 18:57:55 UTC from IEEE Xplore. Restrictions apply.
 TABLE I
T HE PROPOSED GOVERNANCE CRITERIA FOR REMOTE MONITORING AND MANAGEMENT SYSTEM WITH THEIR COMPUTED SCORE AND WEIGHT FOR
W I -F I ACCESS POINTS USE CASE .
Score Weight Criteria
88 100 Did the service provider have declared features or attributes that service provider can monitor or configure by its remote management platform? And is the declaration accessible to users and certificate issuing institutes?
100 100 Did the service provider declare the purpose of remote monitoring and management?
80 100 Did the monitoring system minimize collecting any information relating to an identifiable individual or device such as MAC address of a device or the personal number of individuals?
90 100 Does the monitoring system collect minimized amount of performance parameters of services or analyze data packets as well?
80 100 What interval does the monitoring system collect data from users device or service? (every second, every minute, every hour, every 12 hours, every day)
89 100 Did the service provider declare the location of data storage and level of law enforcement applied? And is the declaration accessible to users and certificate issuing institutes?
90 70 Did the service provider declare data centers location (country or region) that are responsible of monitoring and configuration of devices or services?
100 70 Does the service provider store data on in-house data center or it stores monitored data on the third-party infrastructure?
100 100 Did the service provider declare what level of law enforcement applies to collected data in each data storage location?
80 100 Did the service provider declare data backup and destruction plan for all storage location?
70 80 Does the service provider use latest data protection methods in the industry (e.g. encryption mechanisms) while storing monitored data in its data center?
73 100 Did the service provider declare who has access to the monitored data or can configure service or device? And is the declaration accessible to users and certificate issuing institutes?
70 100 Does the service provider use an industry-grade access control mechanisms e.g. attribute-based access control (ABAC) to control access? (in-house business, in-house operation or third-party access control)
80 80 Does the service provider records data process activities including collecting data and configuration performed automatically or by the service provider?
70 100 Did the service provider minimize configuring critical features or attributes that may make the remote device or service vulnerable?
65 80 Does the service provider employ the latest security and privacy measures for remote monitoring and management system?
80 100 Does the remote monitoring and management system use latest transmission encryption methods in the industry?
50 60 Does the remote monitoring and management platform perform multi-factor cloud authentication?
50 70 Does the monitoring and configuration platform use PKI to secure devices and communication?

weighted data formula as follows: trustworthiness for everyone regardless of expertise. In case of
 the Wi-Fi access point use case, the scores and weights result
 n
 X 2 Wi in overall trustability score of 80, which it means the trustee
Score = ( in )
i i Wi can receive label B according to labeling classification.The
labeling program facilitates decision making for trustors in
The multi-metric method can quantify governance of com- different situations rather than simple yes/no decision.
plex systems in form of system of systems methodology,
which can provide a granular assessment of remote monitoring TABLE II
and management systems. In the multi-metric method, each M APPING TRUSTABILITY SCORES TO LABELS AND THEIR COLOR
REPRESENTATION
criterion gets a score in the 0-100 range, which indicates how
much each criterion accomplished. Besides, weights indicate Score Label Color
the importance of each criterion in the equation such that one 90-100 A
80-90 B
criterion may radically affect the overall trust formation, while 70-80 C
another may have low impact on building trust. 60-70 D
Table I presents scores and weights of governance criteria 50-60 E
<50 F
for trust assessment of Wi-Fi access point monitoring and
management use case. In IoT monitoring and management,
retrieving sensor’s data is the main goal of monitoring so VI. D ISCUSSION
that weight of corresponding criterion asking whether the The certificate or labeling program as the tangible outcome
monitoring system records data packets is low, whilst in the of proposed trust model motivates trustees to perform trust-
mobile application services, the weight will be high value due worthy behaviors and actions, besides facilitates the decision
to high risk of data misuse. Indeed, weights are selected based making for trustors. In addition, representing labels with
on the impact of criteria in trustworthiness of the system, color scheme simplifies identifying the level of compliance
while each criterion has different impact on trustworthiness. to regulation and policies for everyone.
Substituting the use case scores and weights of governance However, organizing an institutional assurance is a challeng-
sub-criteria of each category into the multi-metric equation ing task and requires governmental and key business actors’
will result in scores of each category presented in Table I. support. For example, European GDPR can be a cornerstone
Then, it will require another multi-metric operation in order to provide privacy governance procedures for technology
to calculate overall trustability score of the system, which standardization. On the other hand, definition and updating
results in overall trustability score of 80 for remote monitoring governance criteria requires active participation of governance
and management approach Wi-Fi access point monitoring and bodies and business actors in the same way as technology
management use case presented in Table I. standardization groups. In this respect, Consensus Assess-
The next step will map the measured score to respective ments Initiative Questionnaire (CAIQ) [20] provided by Cloud
certificate or label. The certificate or labeling program has to Security Alliance (CSA) can be a cornerstone for definition
determine limits of labels or certificate threshold per domain. of governance criteria. In addition, the implementation of the
For example, in case of issuing certificate, any trustee that multi-metric method requires determining the importance of
gets a score over 60 can be qualified to receive the trustwor- each criterion. However, the importance of each criterion in
thiness certificate. In the labeling program, trustees will be the multi-metric method can vary among actors per domain
qualified to receive trustworthiness labels according to Table so that machine learning techniques such as neural networks
II classification. This classification can be accompanied with can facilitate weighting process and accomplish a point of
coloring in order to facilitate the understanding of level of agreement among different actors and experts.

217

Authorized licensed use limited to: ULAKBIM UASL - GAZI UNIV. Downloaded on April 25,2024 at 18:57:55 UTC from IEEE Xplore. Restrictions apply.
 VII. R ELATED W ORK and hybrid trust models.
The most of literature utilized the reputation-based approach
In recent years, trust models and trust assessment have been in order to evaluate trustworthiness. However, considering only
investigated in various areas such as wireless sensor networks reputation of a service provider does not imply that services
(WSN), IoT, cloud services and enterprise systems in order to and products are trustable because service providers’ remote
improve reliability of systems. monitoring and management approaches may not comply with
In WSN domain, trust models have been recommended as security and privacy best practices. Hence, compliance to
an effective mechanism to secure WSN, which extensively regulations and best practices can ensure trustability of remote
considers communication behavior in order to evaluate trust. monitoring and management of devices in cloud environment.
The [21] classified trust model into centralized, distributed and In addition, implementing trust models in the machine-to-
hybrid in regard to where trust information stored. Authors in machine communication can not ensure that services are
[22] classified trust models in WSN into node and data trust trustworthy and users will trust to service providers. Hence,
models. In node trust models, nodes calculate trust values in assessing trustworthiness of service provider’s remote mon-
order to be able to associate with each other and node trust itoring and management approach can complement device-
models can be classified into centralized and distributed. In level trust models and accordingly ensure acceptability and
data trust model, nodes calculate trust values to be able to trustability of novel services.
distinguish data of legal nodes from illegal nodes. Trust models
that targeted WSN can also be classified to reputation-based VIII. C ONCLUSION
and credential-based model as well. The [23]–[27] presented
reputation-based trust models for WSN so that nodes can reli- The remote monitoring and management systems are de-
ably associate with each other. The [28] presented distributed signed to improve efficiency and performance, but they need
trust model, which calculated trust according to different as- to be trustworthy so that end-users can accept managed devices
pects including direct trust, recommendation, communication, and services. This paper presented existing trust issues in re-
energy and data trust in order to evaluate trustworthiness of mote monitoring and management and accordingly introduced
nodes in WSN. a trust model for maximizing trustability of remote monitoring
In IoT domain, trust models utilized reputation of devices in and management systems. The proposed trust model employed
order to evaluate trustworthiness. The [29] presented a recom- the multi-metric method to quantify governance criteria de-
mendation and reputation trust model for social IoT devices fined for trustability assessment into trustability scores in
so that IoT devices associate with each other in trustworthy order to generate certificate or labels for remote monitoring
manner. The [30] presented a trust model based on knowledge, and management approaches. The introduced trust model and
experience and reputation trust metrics in order to assess IoT labeling technique can motivate service providers to improve
devices trustworthiness in a network. The [31] presented a trustability of their monitoring and management approaches,
trust framework for IoT devices, in which IoT devices use meanwhile it simplifies trust formation decision making for
public key in order to ensure trust. The [32] presented IoT end-users. Indeed, the proposed trust model paves the way
trust and reputation model by using distributed probabilistic for building trustable remote monitoring and management
neural networks in order to distinguish trustworthy nodes systems. In effect, trustable remote monitoring and manage-
from malicious nodes. The [33] presented a challenge-response ment systems will improve acceptability of new managed
trust assessment for personal space IoT in order to evaluate services and devices, in which enhance performance and user
trustworthiness of the IoT devices before their association to satisfaction.
the personal space.
In cloud environment, trust models utilize different ap- A. Future Work
proaches in order to assess security aspect of trust in cloud The proposed trust model illustrated actors and interactions
services. The [34] classified cloud trust models, which are of a monitoring and management system as well as the essence
based on customer feedback into applied technology and re- of an institutional-based trust assessment. The follow-up work
search models. The applied technology trust models comprises will therefore be developing a labeling program application
policy and SLA negotiation models in order to control access by considering all available regulation for security, privacy,
to cloud and establish trust with customers. The research trust trustability and safety to build a comprehensive governance
models provide approaches to establish trust in cloud service criteria data set. In addition, the development of labeling
environments using weighted average, probability, fuzzy logic, program will consider machine-learning techniques in order
statistical analysis and machine learning methods. The [35] to automate weighting process and simplify quantifying the
presented a trust model to measure security strength and governance criteria for trustability assessment.
calculate a trust value for only security aspect of cloud
services. R EFERENCES
In enterprise information systems, trust models almost
consider security policies of enterprise. The [36] classified [1] IHS Markit. (2018) The internet of things: a movement, not a market.
[retrieved: Sep, 2018]. [Online]. Available: https://cdn.ihs.com/www/
enterprise trust models into credential-based, reputation-based pdf/IoT ebook.pdf

218

Authorized licensed use limited to: ULAKBIM UASL - GAZI UNIV. Downloaded on April 25,2024 at 18:57:55 UTC from IEEE Xplore. Restrictions apply.
 [2] J. M. Morgan, S. Kitt, J. Gill, J. M. McComb, G. A. Ng, [21] V. U. Rani and K. S. Sundaram, “Review of trust models in wireless
J. Raftery, P. Roderick, A. Seed, S. G. Williams, K. K. Witte, sensor networks,” Int. J. Comput. Inf. Syst. Control Eng, vol. 8, pp.
D. J. Wright, S. Harris, and M. R. Cowie, “Remote management of 371–377, 2014.
heart failure using implantable electronic devices,” European Heart [22] G. Han, J. Jiang, L. Shu, J. Niu, and H.-C. Chao, “Management and
Journal, vol. 38, no. 30, pp. 2352–2360, 2017. [Online]. Available: applications of trust in wireless sensor networks: A survey,” Journal
+http://dx.doi.org/10.1093/eurheartj/ehx227 of Computer and System Sciences, vol. 80, no. 3, pp. 602 – 617,
[3] G. Boriani, A. Da Costa, A. Quesada, R. P. Ricci, S. Favale, G. Boscolo, 2014, special Issue on Wireless Network Intrusion. [Online]. Available:
N. Clementy, V. Amori, L. Mangoni di S. Stefano, H. Burri, and http://www.sciencedirect.com/science/article/pii/S0022000013001232
on behalf of the MORE-CARE Study Investigators, “Effects of remote [23] S. Ganeriwal, L. K. Balzano, and M. B. Srivastava, “Reputation-based
monitoring on clinical outcomes and use of healthcare resources in framework for high integrity sensor networks,” ACM Trans. Sen.
heart failure patients with biventricular defibrillators: results of the Netw., vol. 4, no. 3, pp. 15:1–15:37, Jun. 2008. [Online]. Available:
more-care multicentre randomized controlled trial,” European Journal http://doi.acm.org/10.1145/1362542.1362546
of Heart Failure, vol. 19, no. 3, pp. 416–425, 2017. [Online]. Available: [24] F. Gómez Mármol and G. Mart´(i) nez Pérez, “Providing trust in wireless
http://dx.doi.org/10.1002/ejhf.626 sensor networks using abio-inspiredtechnique,” Telecommunication
[4] N. Kumar, J. Madhuri, and M. ChanneGowda, “Review on security and Systems, vol. 46, no. 2, pp. 163–180, Feb 2011. [Online]. Available:
privacy concerns in internet of things,” in 2017 International Conference https://doi.org/10.1007/s11235-010-9281-7
on IoT and Application (ICIOT), May 2017, pp. 1–5. [25] S. Ozdemir, “Functional reputation based reliable data aggregation and
[5] V. Beltran, J. A. Martinez, and A. F. Skarmeta, “User-centric access transmission for wireless sensor networks,” Computer Communications,
control for efficient security in smart cities,” in 2017 Global Internet of vol. 31, no. 17, pp. 3941 – 3953, 2008. [Online]. Available:
Things Summit (GIoTS), June 2017, pp. 1–6. http://www.sciencedirect.com/science/article/pii/S0140366408004301
[6] W. Sherchan, S. Nepal, and C. Paris, “A survey of trust in social [26] J. Zhang, R. Shankaran, M. A. Orgun, V. Varadharajan, and A. Sattar,
networks,” ACM Comput. Surv., vol. 45, no. 4, pp. 47:1–47:33, Aug. “A dynamic trust establishment and management framework for wireless
2013. [Online]. Available: http://doi.acm.org/10.1145/2501654.2501661 sensor networks,” in 2010 IEEE/IFIP International Conference on
[7] L. Mui, “Computational models of trust and reputation : agents, Embedded and Ubiquitous Computing, Dec 2010, pp. 484–491.
evolutionary games, and social networks,” Ph.D. dissertation, [27] S. Misra and A. Vaish, “Reputation-based role assignment for
Massachusetts Institute of Technology, Massachusetts, 2003. [Online]. role-based access control in wireless sensor networks,” Computer
Available: http://hdl.handle.net/1721.1/87343 Communications, vol. 34, no. 3, pp. 281 – 294, 2011, special
[8] T. W. Um, G. M. Lee, and J. K. Choi, “Strengthening trust in the Issue of Computer Communications on Information and Future
future social-cyber-physical infrastructure: an itu-t perspective,” IEEE Communication Security. [Online]. Available: http://www.sciencedirect.
Communications Magazine, vol. 54, no. 9, pp. 36–42, September 2016. com/science/article/pii/S0140366410000885
[28] J. Jiang, G. Han, F. Wang, L. Shu, and M. Guizani, “An efficient
[9] T. Grandison and M. Sloman, Specifying and Analysing Trust for
distributed trust model for wireless sensor networks,” IEEE Transactions
Internet Applications. Boston, MA: Springer US, 2003, pp. 145–157.
on Parallel and Distributed Systems, vol. 26, no. 5, pp. 1228–1237, May
[Online]. Available: https://doi.org/10.1007/978-0-387-35617-4 10
2015.
[10] J. Riegelsberger, M. A. Sasse, and J. D. McCarthy, “The mechanics [29] U. Jayasinghe, N. B. Truong, G. M. Lee, and T. W. Um, “Rpr: A
of trust: A framework for research and design,” International Journal trust computation model for social internet of things,” in 2016 Intl
of Human-Computer Studies, vol. 62, no. 3, pp. 381 – 422, IEEE Conferences on Ubiquitous Intelligence Computing, Advanced and
2005. [Online]. Available: http://www.sciencedirect.com/science/article/ Trusted Computing, Scalable Computing and Communications, Cloud
pii/S1071581905000121 and Big Data Computing, Internet of People, and Smart World Congress
[11] N. Karthik and V. R. S. Dhulipala, “Trust calculation in wireless (UIC/ATC/ScalCom/CBDCom/IoP/SmartWorld), July 2016, pp. 930–
sensor networks,” in 2011 3rd International Conference on Electronics 937.
Computer Technology, vol. 4, April 2011, pp. 376–380. [30] U. Jayasinghe, G. M. Lee, T. W. Um, and Q. Shi, “Machine learning
[12] H. Hu, R. Lu, and Z. Zhang, “Vtrust: A robust trust framework for relay based trust computational model for iot services,” IEEE Transactions on
selection in hybrid vehicular communications,” in 2015 IEEE Global Sustainable Computing, pp. 1–1, 2018.
Communications Conference (GLOBECOM), Dec 2015, pp. 1–6. [31] K. A. R. Rehiman and S. Veni, “A trust management model for sensor
[13] R. K. L. Ko, P. Jagadpramana, M. Mowbray, S. Pearson, M. Kirchberg, enabled mobile devices in iot,” in 2017 International Conference on
Q. Liang, and B. S. Lee, “Trustcloud: A framework for accountability I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC), Feb
and trust in cloud computing,” in 2011 IEEE World Congress on 2017, pp. 807–810.
Services, July 2011, pp. 584–588. [32] S. Asiri and A. Miri, “An iot trust and reputation model based on
[14] Z. Yan, P. Zhang, and A. V. Vasilakos, “A security and trust framework recommender systems,” in 2016 14th Annual Conference on Privacy,
for virtualized networks and software-defined networking,” Security and Security and Trust (PST), Dec 2016, pp. 561–568.
Communication Networks, vol. 9, no. 16, pp. 3059–3069, 2016, sCN- [33] T. Nguyen, D. Hoang, and A. Seneviratne, “Challenge-response trust
14-0760.R1. [Online]. Available: http://dx.doi.org/10.1002/sec.1243 assessment model for personal space iot,” in 2016 IEEE International
[15] Y. Zhang, B. Ge, X. Li, B. Shi, and B. Li, “Controlling a car through Conference on Pervasive Computing and Communication Workshops
obd injection,” in 2016 IEEE 3rd International Conference on Cyber (PerCom Workshops), March 2016, pp. 1–6.
Security and Cloud Computing (CSCloud), June 2016, pp. 26–29. [34] E. F. Rawashdeh, I. I. Abuqaddom, and A. A. Hudaib, “Trust models
[16] QA Cafe. (2017) Is your tr-069 implementation vulnerable to code for services in cloud environment: A survey,” in 2018 9th International
injection attacks? [retrieved: Sep, 2018]. [Online]. Available: https: Conference on Information and Communication Systems (ICICS), April
//www.qacafe.com/training/2017-01-12-tr-069-code-injection-attack/ 2018, pp. 175–180.
[17] European Parliament. (2016) The european data protection regulation. [35] R. Shaikh and M. Sasikumar, “Trust model for measuring security
[retrieved: Sep, 2018]. [Online]. Available: https://eur-lex.europa.eu/ strength of cloud computing service,” Procedia Computer Science,
legal-content/EN/ALL/?uri=uriserv:OJ.L .2016.119.01.0001.01.ENG vol. 45, pp. 380 – 389, 2015, international Conference on
[18] I. Garitano, S. Fayyad, and J. Noll, “Multi-metrics approach for security, Advanced Computing Technologies and Applications (ICACTA).
privacy and dependability in embedded systems,” Wireless Personal [Online]. Available: http://www.sciencedirect.com/science/article/pii/
Communications, vol. 81, no. 4, pp. 1359–1376, 2015. S1877050915004081
[19] A. Fiaschetti, Measurable and Composable Security, Privacy, and [36] A. Manna, A. Sengupta, and C. Mazumdar, “A survey of trust models for
Dependability: The Shield Methodology. Taylor & Francis, a CRC title, enterprise information systems,” Procedia Computer Science, vol. 85,
part of the Taylor & Francis imprint, a member of the Taylor & Francis pp. 527 – 534, 2016, international Conference on Computational
Group, the academic division of T&F Informa, plc, 2017. [Online]. Modelling and Security (CMS 2016). [Online]. Available: http:
Available: https://books.google.no/books?id=fn3ZAQAACAAJ //www.sciencedirect.com/science/article/pii/S1877050916305609
[20] Cloud Security Alliance (CSA). (2017) Consensus as-
sessments initiative questionnaire (caiq). [retrieved: Oct,
2018]. [Online]. Available: https://cloudsecurityalliance.org/artifacts/
consensus-assessments-initiative-questionnaire-v3-0-1/

219

Authorized licensed use limited to: ULAKBIM UASL - GAZI UNIV. Downloaded on April 25,2024 at 18:57:55 UTC from IEEE Xplore. Restrictions apply.